Haywood Kelly, CFA, is vice president of equity research at Morningstar. He'd love to hear from you, and promises to read all your e-mail (even if he can't respond to it all).
Suppose I open a sci-fi restaurant, complete with waiters dressed as aliens, video games, funky Star Trek décor, and so on. (Don't laugh: At a Chicago investment conference last year, a guy was trying to raise money--$10,000 a pop--for just such a loony idea.) The money I spend building the joint is my capital. Whether my business is a good investment depends on how much profit I make as a percentage of that capital. If I earn a profit of $10,000 in a given year, and I've invested $100,000 in building my restaurant, I've just made a return on my capital of 10%. Not spectacular, but better than a savings account.
The way we usually measure return on capital for companies is return on equity (ROE). Dell Computer
earned an incredible 73% on its equity capital--the amount invested in the company by shareholders--in 1998. In other words, for every $1 of shareholder money invested in the firm, Dell generated an annual profit of $0.73. Be careful, though: It's easier to post a large ROE in a single year than it is to do so over a longer period. IBM
, for example, has earned 30% on its equity over the trailing 12 months, but if you average the company's ROEs over the past five years, the figure drops to a much less impressive 11%. It's that long-term return on capital we're interested in.
What about those darlings of '98, Internet stocks? Few Internet companies generate any profits, and aside from American Online
and its 15% ROE, none that I know of generate a high return on capital. When the books are closed on 1998, the return on equity of Amazon.com
is likely to be somewhere around negative 50%. In other words, of the money shareholders have invested in the company, Amazon will have lost half of it in 1998. (To replenish the lost capital, it must turn to shareholders for more money.)
Go back to the restaurant example. Instead of one sci-fi restaurant, let's say I want to open a whole chain of them. (Again, don't laugh. It's as good a business model as that of Rainforest Café
.) In the early years of my empire building, I'll be adding to my capital base aggressively. But the costs of opening restaurants will probably mean I'm making losses; most companies in their formative stages lose money. If after a few years I've sunk $500,000 into my restaurants, but am losing $50,000 annually, my return on capital is negative 10%. (A pretty realistic figure for the sci-fi idea, I'd venture to say.)
It's not necessarily bad for a company to earn a negative return on equity--if it can earn a high return in the future. I'll stomach a negative 10% ROE for my sci-fi restaurants if, in the future, I believe they can earn much higher returns.
The trouble is, in a company's rapid-growth phase, when returns on equity are most often small or negative, it's tough to separate a good business (one that can earn a high ROE) from a bad one (one not able to). After all, each is losing money. Analyzing such companies means asking questions like: Is this a company with enough pricing power to eventually command a premium price for its product? And: Is this a company with enough of a cost advantage that it can undercut the competition? It means, in other words, asking whether the company's business is one that can either generate a high net margin (profit/sales) or a high asset turnover (sales/assets), the two key components of a high return on capital.
Those companies that go on to earn good returns on capital--ROEs of more than, say, 15% or 20%--will probably make good investments. Those that struggle to earn a decent return will probably be wretched investments, regardless of how fast they grow. So if someone tries to talk you into investing $10,000 in a sci-fi restaurant, or in a few hundred shares of an Internet stock, don't ask how fast the company will grow. Ask how the heck it's going to earn a good return on its capital.